I wanted to share my free Blender add-on for Stable Diffusion, which is currently on version 0.0.8. The primary focus of the add-on is making Stable Diffusion another tool for artists.
It has a bunch of features to make this possible:
Seamless texture generation
Cycles render pass to transform renders
Built-in AI upscaling
Inpainting to fix images, or make existing images seamless
Outpainting to extend an image beyond its original size
Is it possible to train the AI on a set of custom images for a specific subject?
I’m curious - can images be effectively saved as 16bit or will it not make a difference to 8bit? I suspect 16b would require more memory…
This is a great add-on btw! I can’t wait until we have a full output for normal maps and other things included! (No pressure, just excited about the prospect)